Yang-Lang Chang

High Performance Computing for Hyperspectral Remote Sensing

Advances in sensor and computer technology are revolutionizing the way remotely sensed data is collected, managed and analyzed. In particular, many current and future applications of remote sensing in Earth science, space science, and soon in exploration science will require real- or near real-time processing capabilities. In recent years, several efforts have been directed towards the incorporation of high-performance computing (HPC) models to remote sensing missions. A relevant example of a remote sensing application in which the use of HPC technologies (such as parallel and distributed computing) is becoming essential is hyperspectral remote sensing, in which an imaging spectrometer collects hundreds or even thousands of measurements (at multiple wavelength channels) for the same area on the surface of the Earth. In this paper, we review recent developments in the application of HPC techniques to hyperspectral imaging problems, with particular emphasis on commodity architectures such as clusters, heterogeneous networks of computers, and specialized hardware devices such as field programmable gate arrays (FPGAs) and commodity graphic processing units (GPUs). A quantitative comparison across these architectures is given by analyzing performance results of different parallel implementations of the same hyperspectral unmixing chain, delivering a snapshot of the state-of-the-art in this area and a thoughtful perspective on the potential and emerging challenges of applying HPC paradigms to hyperspectral remote sensing problems.

Multi-modal gray-level histogram modeling and decomposition

Abstract In this paper, we present a novel multi-modal histogram thresholding method in which no a priori knowledge about the number of clusters to be extracted is needed. The proposed method combines regularization and statistical approaches. By converting the approaching histogram thresholding problem to the mixture Gaussian density modeling problem, threshold values can be estimated precisely according to the parameters belonging to each contiguous cluster. Computational complexity has been greatly reduced since our method does not employ conventional iterative parameter refinement. Instead, an optimal parameter estimation interval was defined before the estimation procedure. This predefined optimal estimation interval reduces time consumption while other histogram decomposition based methods search all feature space to locate an estimation interval for each candidate cluster. Experimental results with both simulated data and real images demonstrate the robustness of our method.

Efficient matching of large-size histograms

As we know, histogram matching is a commonly-adopted technique in the applications of pattern recognition. The matching of two patterns can be accomplished by matching their corresponding histograms. In general, the number of features and the resolution of each feature will determine the size of histogram. The more the number of features and the higher the resolution of each feature, the stronger the discrimination capability of histogram will be. Unfortunately, the increase of histogram size will lead to the decrease of the efficiency of histogram matching because traditional algorithms in evaluating similarity are all relevant to the histogram size. In this paper, a novel histogram-matching algorithm is proposed whose efficiency is irrelevant to the histogram size. The proposed algorithm can be applied to commonly-adopted histogram similarity measurement functions, such as histogram intersection function, L1 norm, L2 norm, χ2 test and so on. By adopting our proposed algorithm, future researchers can focus more on the selection and combination of histogram features and freely adjust the resolution of each feature without worrying the decrease of retrieval efficiency.

Hyperspectral Image Classification Using Nearest Feature Line Embedding Approach

Eigenspace projection methods are widely used for feature extraction from hyperspectral images (HSI) for the classification of land cover. Projection transformation is used to reduce higher dimensional feature vectors to lower dimensional vectors for more accurate classification of land cover types. In this paper, a nearest feature line embedding (NFLE) transformation is proposed for the dimension reduction (DR) of an HSI. The NFL measurement is embedded in the transformation during the discriminant analysis phase, instead of the matching phase. Three factors, including class separability, neighborhood structure preservation, and NFL measurement, are considered simultaneously to determine an effective and discriminating transformation in the eigenspaces for land cover classification. Three state-of-the-art classifiers, the nearest-neighbor, support vector machine, and NFL classifiers, were used to classify the reduced features. The proposed NFLE transformation is compared with different feature extraction approaches and evaluated using two benchmark data sets, the MASTER set at Au-Ku and the AVIRIS set at Northwest Tippecanoe County. The experimental results demonstrate that the NFLE approach is effective for DR in land cover classification in the field of Earth remote sensing.

Greedy modular eigenspaces and positive Boolean function for supervised hyperspectral image classification

This paper presents a new supervised classification tech- nique for hyperspectral imagery, which consists of two algorithms, re- ferred to as the greedy modular eigenspace (GME) and the positive Boolean function (PBF). The GME makes use of the data correlation matrix to reorder spectral bands from which a group of feature eigen- spaces can be generated to reduce dimensionality. It can be imple- mented as a feature extractor to generate a particular feature eigens- pace for each of the material classes present in hyperspectral data. The residual reconstruction errors (RREs) are then calculated by projecting the samples into different individual GME-generated modular eigens- paces. The PBF is a stack filter built by using the binary RRE as classi- fier parameters for supervised training. It implements theminimum clas- sification error (MCE) as a criterion so as to improve classification performance. Experimental results demonstrate that the proposed GME feature extractor suits the nonlinear PBF-based multiclass classifier well for classification preprocessing. Compared to the conventionalprincipal components analysis (PCA), it not only significantly increases the accu- racy of image classification but also dramatically improves the eigende- composition computational complexity.

Data fusion of hyperspectral and SAR images

A novel technique is proposed for data fusion of earth remote sensing. The method is developed for land cover classification based on fusion of remote sensing images of the same scene collected from multiple sources. It presents a framework for fusion of multisource remote sensing images, which consists of two algorithms, referred to as the greedy modular eigenspace (GME) and the feature scale uniformity transformation (FSUT). The GME method is designed to extract features by a simple and efficient GME feature module, while the FSUT is performed to fuse most correlated features from different data sources. Finally, an optimal positive Boolean function based multiclass classifier is further developed for classification. It utilizes the positive and negative sample learning ability of the minimum classification error criteria to improve classification accuracy. The performance of the proposed method is evaluated by fusing MODIS/ASTER airborne simulator (MASTER) images and the airborne synthetic aperture radar (SAR) images for land cover classification during the PacRim II campaign. Experimental results demonstrate that the proposed fusion approach is an effective method for land cover classification in earth remote sensing, and improves the precision of image classification significantly compared to conventional single source classification.

A region-based GLRT detection of oil spills in SAR images

In the study, we propose a fast region-based method for the detection of oil spills in SAR images. The proposed method combines the image segmentation technique and conventional detection theory to improve the accuracy of oil spills detection. From the image statistical characteristics, we first segment the image into regions by using moment preserving method. Then, to get a more integrated segmentation result, we adopt N-nearest-neighbor rule to merge the image regions according to their spatial correlation. Performing the split and merge procedure, we can partition the image into oil-polluted and sea reflection regions, respectively. Based on the segmentation results, we build data models of oil spills and approximate them by using normal distributions. Employing the built oil spills model and the generalized likelihood ratio test (GLRT) detection theory, we derive a closed form solution for oil spills detection. Our proposed method possesses a smaller variance and can reduce the confusion interval in decision. Moreover, we adopt the sample average of image region to reduce the computation complexity. The false alarm rate and oil spills detection probability of the proposed method are derived theoretically. Under the criterion of constant false alarm ratio (CFAR), we determine the threshold of the decision rule automatically. Simulation results performed on ERS2-SAR images have demonstrated the efficiency of the proposed approach.

Foreword to the Special Issue on High Performance Computing in Earth Observation and Remote Sensing

The 17 papers in this special issue present the current state-of-the-art and the most recent developments in the task of incorporating high performance computing techniques and practices to Earth observation and remote sensing problems.

A Parallel Simulated Annealing Approach to Band Selection for High-Dimensional Remote Sensing Images

In this paper a parallel band selection approach, referred to as parallel simulated annealing band selection (PSABS), is presented for high-dimensional remote sensing images. The approach is based on the simulated annealing band selection (SABS) scheme which is originally designed to group highly correlated hyperspectral bands into a smaller subset of modules regardless of the original order in terms of wavelengths. SABS selects sets of correlated hyperspectral bands based on simulated annealing (SA) algorithm and utilizes the inherent separability of different classes to reduce dimensionality. In order to be effective, the proposed PSABS is introduced to improve the computational performance by using parallel computing technique. It allows multiple Markov chains (MMC) to be traced simultaneously and fully utilizes the parallelism of SABS to create a set of SABS modules on each parallel node. Two parallel implementations, namely the message passing interface (MPI) cluster-based library and the open multi-processing (OpenMP) multicore-based application programming interface, are applied to three different MMC techniques: non-interacting MMC, periodic exchange MMC and asynchronous MMC for evaluation. The effectiveness of the proposed PSABS is evaluated by NASA MODIS/ASTER (MASTER) airborne simulator data sets and airborne synthetic aperture radar (SAR) images for land cover classification during the Pacrim II campaign in the experiments. The results demonstrated that the MMC techniques of PSABS can significantly improve the computational performance and provide a more reliable quality of solution compared to the original SABS method.

Band selection for hyperspectral images based on parallel particle swarm optimization schemes

Greedy modular eigenspaces (GME) has been developed for the band selection of hyperspectral images (HSI). GME attempts to greedily select uncorrelated feature sets from HSI. Unfortunately, GME is hard to find the optimal set by greedy operations except by exhaustive iterations. The long execution time has been the major drawback in practice. Accordingly, finding an optimal (or near-optimal) solution is very expensive. In this study we present a novel parallel mechanism, referred to as parallel particle swarm optimization (PPSO) band selection, to overcome this disadvantage. It makes use of a new particle swarm optimization scheme, a well-known method to solve the optimization problems, to develop an effective parallel feature extraction for HSI. The proposed PPSO improves the computational speed by using parallel computing techniques which include the compute unified device architecture (CUDA) of graphics processor unit (GPU), the message passing interface (MPI) and the open multi-processing (OpenMP) applications. These parallel implementations can fully utilize the significant parallelism of proposed PPSO to create a set of near-optimal GME modules on each parallel node. The experimental results demonstrated that PPSO can significantly improve the computational loads and provide a more reliable quality of solution compared to GME. The effectiveness of the proposed PPSO is evaluated by MODIS/ASTER airborne simulator (MASTER) HSI for band selection during the Pacrim II campaign.